The benchmarks compare max rates for communication and steps per second, but is there a good reference for what the MCUs actually need to be capable of in-practice?
For example, with the slowest mcu (16Mhz AVR) when would you start noticing limitations? x32 microstepping? x256? Three steppers? Six? etc.
Is there any way to guess rather than manual testing?
Trying to understand if these benchmarks can be easily translated into real-use performance limitations so I can check if the MCU supports what I want my printer to do.
Benchmark numbers are pretty straightforward.
You know the MAX steps per second.
You know your kinematics.
For Cartesian, it can be approximated to 2 motors spinning simultaneously.
For Delta - 3 motors at toolhead speed.
So, for example, to move CoreXY, X axis at 1000 mm/s both motors should spin at the same speed.
(diagonal movements, coreXY 1 motor x1.44, bed slinger 2 motors at x0.707)
Assume Rotational distance 40 mm.
1000 / 40 * 200 = 5000 full steps per second.
If there are microstepping, like 256.
Simply multiply by 256 → 5000 * 256 = 1280000 per stepper, 2560000 for 2 steppers.
For x16 → 5000 * 16 = 80000 per stepper.
For delta it is complicated, I can only guess that 3 motors should move at the speed equal to the toolhead. So, for the above example, multiply by 3 → 3840000.
Hope that helps.
1 Like
I was thinking it would be more complicated than that. Thank you for the clear examples. 